Mean response and response diversity predict interactive stressor effects on coexistence in competitive communities

Ecosystems increasingly face multiple human-induced stressors. Understanding their joint impact on biodiversity requires examining how intrinsic ecological properties and species responses to stressors affect species coexistence in a community. Here, we use a resource competition model to explore how niche overlap (the extent to which species share resource requirements), mean response (the mean effect of stressors across species) and response diversity (the variation of responses across species to stressors) drive interactive stressor effects on species pairs. Niche overlap had limited effect on stressor interactions. In contrast, when mean response was negative and response diversity low, synergistic effects (neither stressor alone breaks coexistence, but their combination does) were more likely than antagonistic ones (at least one stressor alone breaks coexistence, but the combined effect does not). The opposite held when mean response was positive and response diversity was high. These insights also explained chemical mixture effects in larger communities simulated with a more complex model. We show how chemical concentrations drive mean response and response diversity, which dictate interactive effects. Finally, we apply the theory to empirical data from a field experiment with two environmental stressors. Our results suggest that mean response and response diversity help in understanding interactive stressor effects on coexistence.